How to Read Google PageSpeed Insights and Fix What Matters
Table of Contents
Slow-loading pages cost UK businesses real money. Google’s own research shows that as page load time increases from one second to three seconds, the probability of a mobile visitor leaving rises by 32%. If your web developer has pointed at a red score in Google PageSpeed Insights, or your analytics shows a stubbornly high bounce rate, the two problems are almost certainly connected.
Google PageSpeed Insights is a free performance testing tool. Enter any URL and it returns a composite score from 0 to 100, real user measurements from Chrome browsers, and a prioritised list of recommendations. The score is generated by Lighthouse, an open-source auditing engine maintained by Google. What makes PageSpeed Insights particularly useful is that it combines two data sources most other tools do not: real user measurements from the Chrome User Experience Report (CrUX), and simulated lab data run under controlled conditions.
This guide explains how to read both data sources, what Core Web Vitals actually measure, which numbers Google uses for ranking, and where technical performance improvements deliver the most commercial benefit for SMEs across Northern Ireland, Ireland, and the UK.
What Google PageSpeed Insights Actually Measures
Google PageSpeed Insights produces a performance score between 0 and 100. That number is a weighted composite of several individual metrics, all generated by Lighthouse running a simulated audit on your URL. The simulation uses a mid-range Android device on a throttled 4G connection, which is why mobile scores tend to be lower than desktop scores and why a Lighthouse score of 100 on mobile is genuinely rare.
Understanding the weighting behind the score helps you focus your effort on the right fixes rather than chasing the number itself.
| Metric | Score Weight | What It Measures |
| Total Blocking Time (TBT) | 30% | Time the main thread is blocked from responding to user input |
| Largest Contentful Paint (LCP) | 25% | How quickly the largest visible element loads |
| Cumulative Layout Shift (CLS) | 15% | Visual stability: how much elements shift during load |
| First Contentful Paint (FCP) | 10% | Time until the first text or image is painted |
| Speed Index | 10% | How quickly the page appears visually complete |
| Interaction to Next Paint (INP) | 10% | How fast the page responds after a user interaction |
Colour coding runs consistently through the Google PageSpeed Insights report: green means good, amber means improvement is needed, and red means poor performance. The practical goal is green across all three Core Web Vitals; those thresholds are what Google actually uses as ranking signals, not the composite score.
Decoding Core Web Vitals
Core Web Vitals are the subset of metrics Google has designated as page experience ranking signals. There are currently three: LCP, CLS, and Interaction to Next Paint (INP), which replaced First Input Delay in March 2024. Getting all three into the green band is the practical goal; chasing a high composite score when your Core Web Vitals are already passing is rarely the best use of development time.
Each metric addresses a distinct aspect of how a page feels to a real user, and each has a specific threshold that separates good from needs improvement.
Largest Contentful Paint (LCP): Loading Speed
LCP measures how long it takes for the largest visible element on the page to fully render. For most pages this is a hero image, a prominent heading, or a video thumbnail. Google’s good threshold is under 2.5 seconds.
Common causes of poor LCP on UK websites include unoptimised images served without WebP or AVIF compression, slow server response times on shared hosting without a content delivery network, and render-blocking scripts loaded in the <head> before the page can paint. For businesses on shared hosting, Time to First Byte (TTFB) is often the single biggest LCP lever. Switching to managed WordPress hosting with a London or Dublin edge node can reduce server response time by 400 to 500 milliseconds before a single image is changed, which in many cases is enough to push LCP into the green on its own.
Cumulative Layout Shift (CLS): Visual Stability
CLS measures visual instability. When an image loads without defined dimensions and pushes the navigation down, or a cookie consent banner slides in and shifts your main content, that contributes to your CLS score. Google’s good threshold is below 0.1.
For UK and Irish businesses, CLS is particularly relevant right now. GDPR-compliant cookie consent banners, required under ICO guidance, are one of the most common sources of layout shift for UK sites. If your consent tool is not reserving space in the DOM before it loads, it will degrade your CLS score regardless of how clean the rest of your site is. The fix is straightforward: set a min-height reservation for the banner container in your CSS so the space exists before the JavaScript fires. Most TCF 2.2-compliant platforms (OneTrust, Cookiebot, CookieYes) allow you to adjust the banner’s CSS directly.
Interaction to Next Paint (INP): Responsiveness
INP replaced First Input Delay as a Core Web Vital in March 2024. Where FID only measured the first interaction on a page, INP measures every interaction throughout the session and reports the worst. The good threshold is under 200 milliseconds.
“INP is the first Core Web Vital that reflects how a page genuinely feels to use, not just how quickly it loads,” says Ciaran Connolly, founder of Belfast digital agency ProfileTree. “For SMEs with content-heavy WordPress sites or WooCommerce stores, INP problems are almost always caused by bloated JavaScript: third-party scripts, chat widgets, or page builders firing event listeners on every click. Identifying and deferring those scripts is usually where we start.”
Poor INP typically appears on sites using heavy page builders, sites with multiple third-party tracking scripts loaded synchronously, and e-commerce sites where product pages fire requests on every interaction. For non-developers, the debugging approach is to open Chrome DevTools, go to the Performance tab, click Record, interact with the page, then stop recording. Look for long tasks in the flame chart. Anything blocking the main thread for more than 50 milliseconds is a candidate for optimisation.
Why Your Score and Your Google Ranking Are Not the Same Thing
Understanding how Core Web Vitals translate into ranking signals matters before you spend time on fixes. The composite Lighthouse score and Google’s ranking inputs are not the same thing, and confusing them leads to wasted development effort.
Google uses a pass or fail model for Core Web Vitals in its ranking algorithm. A page either meets the good threshold for LCP, CLS, and INP, or it does not. Moving your LCP from 2.4 seconds to 1.8 seconds, both technically in the green, carries no additional ranking benefit. Chasing a score of 100 when all three Core Web Vitals are already passing is not a useful investment of development time.
The composite score is genuinely useful for three things: benchmarking against competitors, catching regressions after site updates, and identifying which specific technical issues to prioritise. For most SMEs, three problems account for the majority of score loss: unoptimised images, render-blocking third-party scripts, and slow server response times from underspecified hosting.
Google will not penalise you for a score of 85 versus 100. The threshold that matters is the green band on each Core Web Vital. The point where poor performance actually damages commercial outcomes is at the extremes: pages that take five or more seconds to load on mobile, or that shift content visibly as it settles, lose users before they can convert.
For Northern Ireland and Irish businesses, hosting location matters more than many site owners realise. A server based in the United States or Australia adds latency that image compression alone cannot recover. UK-specific hosting with edge nodes in London or Dublin is typically the highest return-on-investment performance change for regionally focused businesses, ahead of any other individual optimisation. Our SEO services explain how ProfileTree handles technical SEO audits alongside ongoing search work for clients.
How to Use Google PageSpeed Insights: A Step-by-Step Walkthrough
Running a PageSpeed Insights audit takes less than a minute, but interpreting the results correctly takes more care. The report is structured in layers, and most users stop at the top-level score when the most actionable data sits further down.
To run an audit: go to pagespeed.web.dev, enter the URL of the page you want to test, and click Analyse. Results take 15 to 30 seconds. Run the test separately for mobile and desktop by using the tabs at the top of the results page. For most UK businesses, the mobile result is the one that matters most for SEO.
Field Data: What Real Users Experience
At the top of the report, below the score, is the Field Data section. This draws from the Chrome User Experience Report, a dataset of real Chrome browser interactions from visitors who have been to your page over the past 28 days. This is the data Google uses for ranking signals, not the Lighthouse simulation.
Field data can look better or worse than the score depending on your actual visitor base. If most of your visitors arrive on fast broadband connections, your field LCP may sit in the green even if the simulated audit returns amber. Conversely, if you serve a predominantly mobile audience in rural areas of Northern Ireland or the west of Ireland, where connectivity varies, your field data may reflect a worse experience than the controlled lab test suggests.
If your page is new or low-traffic, the field data section will show “not enough data.” That is normal. PSI requires a minimum volume of Chrome user visits before it populates, so use lab data as your guide until the field data fills in.
Lighthouse Lab Data: Your Diagnostic Starting Point
Below field data, the Lighthouse lab results show the performance score and individual metric readings. These are affected by the simulated throttling, so treat them as a diagnostic tool rather than a definitive picture of user experience.
The lab data is the most immediately useful section when you are trying to identify what to fix. It updates in real time after changes, unlike field data which has a 28-day lag. If you have deployed a fix, run a fresh PageSpeed Insights test and check the lab data to confirm the change has had the expected effect before waiting four weeks for field data to catch up.
Opportunities and Diagnostics
This is where Google PageSpeed Insights becomes actionable. The Opportunities section lists prioritised recommendations, each with an estimated time saving attached. Google PageSpeed Insights orders these by potential impact, so work from the top down. Common opportunities on UK business websites include:
- Properly sizing images: serving images at a larger resolution than their display dimensions wastes bandwidth and slows LCP
- Eliminating render-blocking resources: CSS or JavaScript in the <head> that delays page painting
- Reducing unused JavaScript: scripts that load but are not used on the page add to TBT
- Serving images in modern formats: WebP and AVIF deliver the same visual quality at 25 to 35% smaller file sizes than JPEG or PNG
- Removing unused CSS: stylesheet bloat that inflates the critical rendering path
Below Opportunities, the Diagnostics section covers checks that do not have direct time savings attached but affect overall performance: whether images have explicit width and height attributes (which prevents CLS), whether fonts are loaded efficiently, and whether third-party scripts are adding excessive execution time.
The Passed Audits section is worth a quick check after any site update. It confirms nothing previously working has been inadvertently broken. For a detailed walkthrough of the errors you are most likely to encounter across both PSI and Search Console together, our guide to the most common Search Console errors and how to fix them covers both tools in context.
PageSpeed Insights and Your Business: Where It Connects to Revenue
When ProfileTree works with a client in Belfast whose WooCommerce store shows a red LCP score, the commercial problem is almost always the same: slow product pages are losing customers before they reach the add-to-cart button. The PageSpeed Insights report identifies the mechanism; the revenue impact is what justifies the fix.
Performance improvements connect directly to three business metrics. The first is bounce rate: users who wait more than three seconds on a slow mobile page are statistically more likely to leave before the page finishes loading. The second is conversion rate: Google and Deloitte research found that a one-second improvement in mobile load time can lift conversions by up to 27% in retail, though results vary by site and audience. The third is Google Ads Quality Score: mobile landing page speed is a direct input into Quality Score for Search campaigns, meaning a slow page raises your cost per click on top of reducing your organic visibility.
For UK businesses running WooCommerce or Shopify stores, this double cost is significant. Every pound spent acquiring traffic through paid or organic channels is partially wasted if the landing page loses users in the first three seconds.
The fixes are not always complex. Switching to a faster image format, moving third-party scripts to load after the page is interactive, and upgrading to UK-based managed hosting with a CDN will address the majority of Core Web Vitals failures on most small business sites. Where the underlying architecture needs deeper work, such as a legacy site built on a slow page builder, the conversation moves into a rebuild or a platform migration.
Our web development service covers technical performance work as part of every build, including Core Web Vitals optimisation and server configuration. For sites with existing performance problems, a website analysis is the logical starting point before deciding whether to optimise the current build or move to a faster platform.
Alternative Speed Audit Tools Worth Knowing
Google PageSpeed Insights is the authoritative source for Core Web Vitals assessment, but it is most useful alongside complementary tools that provide different views of the same performance problems.
GTmetrix uses the same Lighthouse performance metrics, so scores should broadly align with PageSpeed Insights. The key difference is geographic flexibility: GTmetrix lets you choose the server location for the test, which is useful for confirming how your site performs from a UK or Irish node specifically. GTmetrix also includes a Waterfall chart showing every request the page makes with individual timings, which is significantly more useful for diagnosing specific bottlenecks than the PSI Opportunities list alone. Use GTmetrix when you need to understand why something is slow; use PageSpeed Insights when you need to know whether Google considers it acceptable.
Pingdom does not use Lighthouse and does not report Core Web Vitals, making it less relevant for SEO-focused audits today. It provides a straightforward performance grade based on load time, which is useful for a quick sanity check but insufficient for any detailed technical work.
Chrome DevTools (the Performance tab and the Network tab) is the most powerful diagnostic option but requires technical knowledge to interpret. For debugging INP specifically, it is indispensable. The PageSpeed Insights report identifies that an INP problem exists and estimates its severity; DevTools shows you exactly which JavaScript task is causing the delay and on which interaction.
One practical note on interpreting discrepancies: PSI, GTmetrix, and your own browser will sometimes return different scores for the same page at the same moment. This is expected behaviour. The simulated environment, test server location, network conditions, and browser caching state all affect results. A meaningful performance improvement will show up consistently across all tools; a discrepancy of 10 to 15 points between tools on a single test is noise, not signal.
Getting Started with a PageSpeed Audit for Your Site
Running Google PageSpeed Insights is free and takes a few minutes, but knowing what to do with the results is where most site owners get stuck. The report can return 15 or 20 individual recommendations, and without a clear prioritisation framework, it is easy to spend time on fixes that move the score but do not meaningfully improve the user experience or the Core Web Vitals pass or fail status.
A practical starting point: focus only on the red and amber Core Web Vitals in the field data section first. If LCP, CLS, or INP are failing, fixing those has direct ranking implications. Once all three Core Web Vitals are in the green on field data, the remaining score optimisations are about user experience improvement rather than SEO lift, and can be prioritised accordingly.
If you are a business owner rather than a developer, the most useful action after reviewing your PageSpeed Insights report is to share the URL and the report results with your web development team or agency. The Opportunities section is written in developer-readable language; a competent team will be able to estimate the effort and impact for each item and give you a prioritised fix list.
What the Results Mean for Your Business
The prioritisation framework above is useful, but it is worth understanding why Core Web Vitals failures carry a commercial cost beyond just SEO rankings.
When ProfileTree works with a client in Belfast whose WooCommerce store shows a red LCP score, the commercial problem is almost always the same: slow product pages are losing customers before they reach the add-to-cart button. The PageSpeed Insights report identifies the mechanism; the revenue impact is what justifies the fix.
Performance improvements connect directly to three business metrics. The first is bounce rate: users who wait more than three seconds on a slow mobile page are statistically more likely to leave before the page finishes loading. The second is conversion rate: Google and Deloitte research found that a one-second improvement in mobile load time can lift conversions by up to 27% in retail, though results vary by site and audience. The third is Google Ads Quality Score: mobile landing page speed is a direct input into Quality Score for Search campaigns, meaning a slow page raises your cost per click on top of reducing your organic visibility.
For UK businesses running WooCommerce or Shopify stores, this double cost is significant. Every pound spent acquiring traffic through paid or organic channels is partially wasted if the landing page loses users in the first three seconds.
The fixes are not always complex. Switching to a faster image format, moving third-party scripts to load after the page is interactive, and upgrading to UK-based managed hosting with a CDN will address the majority of Core Web Vitals failures on most small business sites. Where the underlying architecture needs deeper work, such as a legacy site built on a slow page builder, the conversation moves into a rebuild or a platform migration.
Our web development service covers technical performance work as part of every build, including Core Web Vitals optimisation and server configuration. For sites with existing performance problems, a website analysis is the logical starting point before deciding whether to optimise the current build or move to a faster platform.
Conclusion
Google PageSpeed Insights gives you a clear, free picture of where your site stands on the metrics Google actually uses. Getting all three Core Web Vitals into the green is the practical goal; everything else is refinement. If your report is showing red or amber on field data and you are not sure where to start, ProfileTree’s technical SEO team works with SMEs across Northern Ireland, Ireland, and the UK to turn those findings into a prioritised fix plan. Get in touch to discuss a Core Web Vitals audit for your site.
Frequently Asked Questions
What is a good Google PageSpeed Insights score?
The composite score is not what Google uses for ranking. The practical goal is green on all three Core Web Vitals in field data: LCP under 2.5 seconds, CLS under 0.1, and INP under 200 milliseconds. A Lighthouse composite score of 75 or above with all three Core Web Vitals passing is a reasonable benchmark for a well-performing SME site. Scores above 90 generally indicate no significant technical performance issues remain.
Does PageSpeed Insights directly affect Google rankings?
Not the composite score. The Core Web Vitals thresholds do. Google uses a pass or fail model: a page either meets the good thresholds for LCP, CLS, and INP or it does not. Pages that fail Core Web Vitals may rank lower in competitive searches where other signals are otherwise equal, but failing a Core Web Vital is not an automatic ranking penalty in the way a manual action or a spam policy violation would be.
Why is my mobile score always lower than my desktop score?
PageSpeed Insights simulates a mid-range Android device on a throttled 4G connection for mobile audits. Desktop tests simulate a faster connection and more processing power. Images served at 2,000 pixels wide that display at 400 pixels on mobile are downloading far more data than needed, adding directly to load time. Responsive image serving, proper size attributes, and modern formats like WebP or AVIF close most of the gap between mobile and desktop scores.
What is the difference between lab data and field data in PageSpeed Insights?
Lab data comes from Lighthouse running a simulated audit in a controlled environment with fixed device and network settings. Field data comes from real Chrome users who have visited your page, recorded over a rolling 28-day window. Field data is what Google uses for ranking signals. Lab data reflects changes immediately after a fix; field data takes up to four weeks to update. If your page lacks sufficient traffic, field data will not populate and you should use lab data as your guide.
What is Interaction to Next Paint and why does it matter?
INP replaced First Input Delay as a Core Web Vital in March 2024. Where FID measured only the delay before the first user interaction was processed, INP measures the response time for every interaction during a page session and reports the worst. It is the most accurate reflection of how responsive a page feels to use. The good threshold is under 200 milliseconds. The most common cause of poor INP is JavaScript running on the main thread in response to user actions, typically from third-party scripts, page builders, or tracking pixels.
How do I fix a CLS problem caused by my cookie banner?
This is one of the most common CLS issues for UK websites under GDPR. Set a min-height on the banner container in your CSS that matches the banner’s rendered height so space is reserved before the JavaScript loads. Most compliant consent platforms (OneTrust, Cookiebot, CookieYes) allow direct CSS adjustment. If the banner uses position: fixed as an overlay, confirm it is not shifting any existing document content on load.
How long does it take to see improvements in field data after fixing PageSpeed issues?
Field data in Google PageSpeed Insights uses a 28-day rolling window from the Chrome User Experience Report. This means fixes you deploy today will take up to four weeks to be fully reflected in field data as the window gradually replaces older measurements with newer ones. Lab data updates immediately, so use that to confirm a fix has worked technically, then monitor field data over the following weeks to see the real-world improvement confirmed.
Can I use PageSpeed Insights to check competitor websites?
Yes. PageSpeed Insights works on any publicly accessible URL that is not blocked by robots.txt or password protection. Checking your closest competitors gives you a baseline comparison for their Core Web Vitals performance and can help you identify whether a speed advantage is available to exploit, particularly if a competitor has known performance problems and you are targeting the same keywords.